Saving lives, spending less - Methodology - World Health Organization
←
→
Page content transcription
If your browser does not render page correctly, please read the page content below
Saving lives, spending less Methodology
Document number: WHO/NMH/NVI/18.9 © World Health Organization 2018. Some rights reserved. This work is available under the CC BY-NC-SA 3.0 IGO licence.
CONTENTS Analytic scope ........................................................................................................................................................ 2 Intervention impact sizes ........................................................................................................................................ 4 Coverage and scale up patterns ............................................................................................................................... 5 Clinical interventions .......................................................................................................................................... 6 Tobacco control interventions ............................................................................................................................ 6 Alcohol reduction policies .................................................................................................................................. 7 Unhealthy Diets .................................................................................................................................................. 8 Physical inactivity mass media ........................................................................................................................... 9 The health benefits of scaling up intervention coverage ....................................................................................... 10 Estimating health care costs of interventions........................................................................................................ 12 Estimating the economic and social returns on investment .................................................................................. 15 Translating avoided deaths into economic returns ............................................................................................ 15 Translating incident cases avoided into economic returns ................................................................................ 15 Social benefits of increased years of healthy life .............................................................................................. 15 Calculating the Return on Investment .............................................................................................................. 15 Summary of Findings ........................................................................................................................................... 16 Costs of scaling up action ................................................................................................................................. 16 Health benefits .................................................................................................................................................. 18 The economic benefits of investment ............................................................................................................... 20 Return on Investment ........................................................................................................................................ 20 Limitations of analytical framework ..................................................................................................................... 22 References ............................................................................................................................................................ 23
ANALYTIC SCOPE The analysis for the economic data underpinning the NCD Global Business Plan was overseen by an informal advisory group consisting of the individuals listed in table 1. The informal advisory group met in November 2017 to review the proposed methodology, scope and preliminary results, and provide direction on the final report. Table 1: Informal Advisory Group members Name Organization Kelly Henning Bloomberg Philanthropies Judith Mackay Vital Strategies Ala Alwan University of WAshington Johanna Ralston World Obesity Federation Dean Jamison University of California, San Francisco Thomas J. Bollyky Council on Foreign Relations This scope of the analysis follows the proposal of the informal advisory group and focuses on the 16 “best buys” for NCDs as described in the Appendix 3 of the Global Action Plan for NCDs, updated in 2017 1 (table 2). The best buys are a set of evidence based priority interventions for NCD prevention and control, which include considerations of cost-effectiveness, feasibility, affordability, acceptability and potential for successful implementation. The analysis covers almost all low and lower middle income countries1. Table 2: WHOs NCD ‘Best Buys’ Interventions included in the analysis TOBACCO Increase excise taxes and prices on tobacco products Enact and enforce comprehensive bans on tobacco advertising, promotion and sponsorship Implement large graphic health warnings on all tobacco packages Eliminate exposure to second-hand tobacco smoke in all indoor workplaces, public places, public transport ALCOHOL Increase excise taxes and prices on alcohol products Enact and enforce bans or comprehensive restrictions on exposure to alcohol advertising (across multiple types of media) Enact and enforce restrictions on the physical availability of retailed alcohol (via reduced hours of sale) 1 This excludes Kosovo and West Bank and Gaza Strip which are not WHO member states. In addition South Sudan, Somalia, Syrian Arab Republic, and the Democratic People’s Republic of Korea have been excluded due to a lack of GDP data available to project prices. Kiribati has been excluded as insufficient baseline epidemiology data was available. 2
UNHEALTHY DIETS Reduce salt intake by engaging the industry in a voluntary reformulation process Reduce salt intake through implementation of front-of-pack labelling Reduce salt intake through a behaviour change communication mass media campaign Reduce salt intake through establishment of a supportive environment in public institutions such as hospitals, schools and nursing homes to enable low sodium meals to be provided PHYSICAL INACTIVITY Increase physical activity rates through a behaviour change communication mass media campaign PHARMACEUTICAL CVD INTERVENTIONS Combination drug therapy post even and for those at 20% or greater risk of CVD event over the coming 10 years CERVICAL CANCER HPV vaccine in 13 year old girls Prevention of cervical cancer through screening and treatment 3
INTERVENTION IMPACT SIZES Two alternative scenarios are measured using the OneHealth Tool impact module – the first in which the intervention coverage scales up to the target coverage levels in either a linear fashion from 2018-2030 or applies policy interventions in the year in which they are enacted, and the counterfactual in which the coverage remains at the baseline level until 2030, assuming no additional investment is made to NCD prevention and treatment. The difference in number of fatal and non-fatal CVD events between the two scenarios represents the health gain attributed to additional investment in an intervention. Table 3: Impact sizes of interventions modelled Intervention Effect Size on tobacco prevalence Comments on evidence Increase excise taxes and prices Elasticity is -0.2 to- 0.52 Based on an assumed tax on tobacco products increase that increases the retail price of cigarettes by 25%. Enact and enforce comprehensive Reduction in prevalence of 10% if bans on tobacco advertising, implemented at the highest intensity promotion and sponsorship level3 Implement large graphic health Reduction in prevalence of 4% if warnings on all tobacco packages graphic health warnings implemented at the highest intensity level3 Additionally, implement plain/standardized packaging Eliminate exposure to second- Reduction in prevalence of 4% if hand tobacco smoke in all indoor implemented at the highest intensity workplaces, public places, public level3 transport Implement effective mass media Reduction in prevalence of 3.8% if campaigns that educate the public implemented at the highest intensity about the harms of level3 smoking/tobacco use and second hand smoke Reduce salt intake by engaging 2.2 g/day salt reduction4 Based on Argentina experience the industry in a voluntary reformulation process Reduce salt intake through 1.8g/day salt reduction men The experience in Finland implementation of front-of-pack indicated that salt intake reduced labelling 1.0 g/day salt reduction women5 by 15% following implementation of a labelling system5. This translates to the gram per day reductions indicated.6 Reduce salt intake through a 5% reduction in salt intake per day7 Movement from 8.48 to 8.05 behaviour change communication g/day via urinary excretion in mass media campaign Vietnam following BCC campaign.7 The same campaign in Australia saw a 10% reduction 4
in sodium intake8. We have taken the conservative option. Reduce salt intake through 7% reduction in salt intake per day A British study on implementing establishment of a supportive standards for school meals shows environment in public institutions a 30% reduction in sodium such as hospitals, schools and intake.9 An Australian study nursing homes to enable low shows a 20% reduction in sodium meals to be provided sodium intake.10 We take the more conservative Australian study as the base, along with the assumption that with one out of three daily meals eaten in public places, the overall impact on daily sodium intake would be one third of that observed in the school lunches Drug therapy (including -1.05 mmol/L change in cholesterol11 Intervention impact is mediated glycaemic control for diabetes via the risk prediction equation12 mellitus and control of 5.9mmHg reduction in systolic blood hypertension using a total risk pressure6 Additional supportive evidence approach and counselling to suggests a reduction in stroke individuals who have had a heart incidence of 80% and IHD attack or stroke and to persons incidence of 88% predicted in with high risk (≥ 30%) of a fatal use of fixed dose polypill13 and non-fatal cardiovascular event in the next 10 years Treatment of new cases of acute acetylsalicylic acid14 myocardial infarction with acetylsalicylic acid, reduction in CVD mortality 15%, ischemic stroke mortality 30%, haemorrhagic stroke mortality 20% COVERAGE AND SCALE UP PATTERNS For the purposes of this analysis, ambitious implementation and coverage scale up patterns were modelled, in order to demonstrate the impact that an increased commitment to NCDs could yield. For all interventions, expert opinion was sought from WHO technical groups related to the quickest possible time in which interventions were believed to be implementable. Table 4 provides a summary of scale up patterns. Table 4: Intervention Scale Up Summary Intervention Scale up time frame Clinical interventions for cardiovascular disease From current to 50% coverage of population in need by and cervical cancer 2030 Tobacco “P” – protect people from tobacco 1 year – begin 2019, implement 2020 smoke Tobacco “W” - warnings -- mass media 1 year – begin 2018, implement 2019 -- packaging 3 years – begin 2018, implement 2021 Tobacco “E” – enforcement of bans on 1 year – begin 2019, implement 2020 5
advertising Tobacco “R” – raise taxes 3 year step-wise pattern to reach 75% of price by 2027-30 Salt “H”– reformulation 3 years to implement, with 28 years to reach full impact potential Salt “A”– front of pack labelling 3 years to implement, with 28 years to reach full impact potential Salt “K” – mass media campaigns 1 year – begin 2018, implement 2019 Salt “E” -- low salt food in public spaces 3 years to implement, with 28 years to reach full impact potential Alcohol – marketing restrictions 2 years, beginning in 2018 Alcohol - taxation 3 year step-wise pattern to reach 50% of price by 2027-30 Alcohol – advertising restrictions 2 years, beginning in 2018 Physical Activity mass media 1-2 years dependent on readiness CLINICAL INTERVENTIONS For the clinical interventions – treatment of those at high risk of cardiovascular disease or post event with multidrug therapy and cervical cancer screening and treatment in the absence of country level data collection systems to monitor coverage, and based on expert knowledge, it was assumed that current coverage in low and lower middle income countries is no higher than 5% of needs being met. In order to reach the SDG target of meeting 50% of current unmet need by 2030, we assumed a linear scale up from the current value to 50% coverage in 2030. The assumption that the health system will strengthen and increase capacity in the appropriate way to meet this target is implicit within the analysis, however as the health system must strengthen as a whole, this has not been costed. For understanding of resource needs for health system strengthening, readers should review the SDG health price tag analysis15 TOBACCO CONTROL INTERVENTIONS The intensity of the MPOWER interventions are graded from 1 to 4, with 1 being the lowest and 4 being the highest level of implementation. These grades correspond to the scores in the most recent WHO Report on the Global Tobacco Epidemic (GTCR)16, with the exception of R (expressed in total tax share), and W-graphic warnings/plain packaging (modified scale that includes plain/standardized packaging at level 5). We assume a scenario, where the scale-up happens ambitiously, and at the soonest possible time for all countries. However, we also take into consideration what is historically possible, meaning that these immediate scale-ups or jumps to the highest level of implementation have been observed before. The baseline figures for these interventions were sourced from GTCR 2017 (2016 data), while scale up was assumed to begin in 2019 – the year after the publication of the NCD business case. 6
For P – Smoke-free policies & E – Tobacco Advertising, Promotion and Sponsorship bans: The consensus was that these are interventions with no specific pattern of scaling up (i.e. a country at level 1 or 2 can just as easily jump to level 4 as those countries in level 3). Also, there is no set time-frame by which these interventions occur. Since these are legislative processes in most countries, the team considered it best to allocate 1 full year for these processes (drafting the law, debates, approval, etc.), before its full implementation can begin. Hence for the ambitious scenario, we assume that 2019 is when the process of changing the policy will begin, with 2020 as the first full year of implementation for all countries at the highest level. For W – Mass Media: Since this is usually an executive function that mainly requires budgetary allocation to be implemented, the consensus of the team was that it could be scaled up immediately given that funds are available. Because of this, all countries scale up by 2019. For W – Graphic Warnings / Plain Packaging: Appendix 3 lists these interventions together, however plain packaging is not included in the GTCR score. To resolve this, it was decided that plain packaging be added as an additional level of implementation (level 5). This makes sense because all countries that have implemented plain packaging are already scored at the highest level in GTCR. In addition, the effect size for plain packaging, although separate, can be added into the projection in the OneHealth Tool. With regards to the scale up pattern, similar to smoke-free policies and TAPS bans, this usually requires some form of legislative approval which means having at least one year allocated for that. Looking at historical data, countries could also jump from having the lowest level of implementation to the highest level. However, the team’s expert opinion is that this particular intervention would need a longer time frame to be implemented, considering that the highest level of implementation (for this simulation) is plain packaging. Hence the earliest time for full implementation would be 2021 among countries that aren’t already at the highest level. For R – Taxation: The team considered the data from GTCR and observed that countries below the highest level of implementation (with less than 75% total tax share of retail price) increased their taxes from 2% to 2.5% per year on average. However, once a country has had a significant increase, it would take at least another 3 years before the next significant increase is implemented. This is the main rationale for the 3-year stepwise scale-up pattern, and the corresponding magnitude of 7.5% for each step (2.5% each year). We are targeting the highest level of implementation for all countries by 2030; therefore the scenario is built so that all countries reach a 75% tax rate by 2028-2030. In order to reach this level, and following the determined scale up pattern of 7.5% every three years, all countries needed to have a tax rate of at least 52.5% by the first step of scaling-up in 2019. This is the reason why countries that had lower tax shares (below 52.5%) have a slightly different pattern. In keeping with the aggressive scale-up, it was decided that the first jump would be to 52.5% for countries below that threshold. Though certainly very ambitious for some countries, the team considered this as a possible scenario given that historical GTCR data shows the maximum tax share jump to be an increase of 48 percentage points, which happened with a country that initially had a 2% tax share. This is also consistent with the observation that a jump of this magnitude is more feasible if coming from a very low starting point, compared to having the same jump from ~30% to 75%. After this first scale-up, it would then revert to the 7.5% increase every three years until it reaches 75% in 2028. For countries already at the highest level of implementation (above 75%), it was observed that taxes still increased by approximately 1% per year. In keeping with the 3-year pattern, these countries would then increase their tax rates 3% at a time, until their rates have surpassed 80%, after which no additional increases are projected. ALCOHOL REDUCTION POLICIES A baseline of coverage was constructed based on the NCD Progress Monitor 17. The following exclusions and clarifications were made to the NCD Progress Monitor data: 7
Following the NCD Progress Monitor, those countries that were marked as ‘n/a’ in the previous NCD Progress Monitor were excluded. This applied to: Bolivia, Haiti, Kyrgyzstan, Micronesia, Nepal, Papua New Guinea, Solomon Islands, and Tunisia. ‘Dry’ countries (i.e. those in which alcohol is prohibited) were included, and the policy recorded as in place. This applied to: Afghanistan, Somalia, Sudan and Yemen. As with the approach taken for the above tobacco control policies, expert input was sought on the quickest feasible scale-up for countries, based on historical precedent. For legislation based alcohol reduction interventions, i.e. reducing the availability of alcoholic products through reduced hours of sale, and restricting advertising, the following approach was taken: Based on historical precedent, it was assumed that the fastest a country could implement a policy from not having a policy in place through development of the legislation to the point of implementation would be two years. The same approach has been used for legislation-based tobacco control interventions. This was then ‘scored’, by creating a dual scenario from the baseline 2018 data; i.e. an enforcement score of 1 was given for countries that did not have a policy in place in 2018, and an enforcement score of 2 was given for those that did. For those countries with an enforcement of 1, costing for the planning, development and then full implementation of the policy was applied, with full implementation costs being reached in 2020. For those countries with an enforcement score of 2, the costs of full implementation of the policy were applied from 2018. For taxation on alcoholic products, the following approach was taken: Baseline data was constructed from the OneHealth Tool, with taxation as a percentage of retail price recorded separately for beer, wine and spirits. A scale- up scenario was then applied, following the tobacco taxation model, in which the tax on all products increases every 3 years, reaching 50% (or above) in 2030. For those countries currently below a level 12.5% tax of retail price, the first increase was to 12.5%, and then 12.5% every three years thereafter, thereby reaching 50% by 20301. For those countries that that already have a taxation level above 50%, the same increase every 3 years was applied, with a ceiling of 200%. These levels of taxation are in keeping with the very wide range of taxation levels for alcoholic products seen across the world. UNHEALTHY DIETS Limited country experiences exist for the SHAKE package of sodium reduction policies. The first assumption made was that a minimum of 3 years is required in order to create the political economy within which sodium reduction policies can be implemented. The fastest scale up is likely to be in mass media campaigns incorporating behaviour change communication techniques which aim to change population behaviour around adding salt at the table. The other interventions, particularly reformulation processes will require some time not only due to the political economy but also due to the time it will take companies to complete reformulation processes. Current country implementation was taken from the NCD progress monitor 17. An assessment was then made by the WHO technical teams about the readiness of countries to implement new policies, and an implementation year ranging from immediate to 2020 was proposed. In 12 high income countries where packages of salt reduction policies have been implemented over a period of time, the average rate of change seen in salt intake is between 1 and 2% per year 18. Figure X below shows tracking over time of these countries salt intake. The maximum change seen is in Finland where a reduction in intake of 38% has been observed over a period of 28 years. The combination of 4 salt reduction policies that are best buys have an impact of approximately 40% reduction in salt intake. Therefore we make the assumption that it will take this 28 year time frame to reach full implementation of these interventions. As this purpose of this modelling exercise was to model ambitious scale up patterns, we assume that 50% of the full possible health benefit could be seen by 2030. Figure 1: historical examples of salt reduction policy impact 8
PHYSICAL INACTIVITY MASS MEDIA All countries were assumed to need a lead time of 1-2 years to develop a mass media campaign, followed by an implementation year during which health benefits begin to be seen. Countries existing policies were taken from the NCD Progress Monitor17 . An assessment was then made by the WHO technical teams about the readiness of countries to implement new policies, and an implementation year ranging from immediate to 2020 was proposed. 9
THE HEALTH BENEFITS OF SCALING UP INTERVENTION COVERAGE We use the NCD impact module of the inter-UN agency OneHealth Tool (OHT) to calculate the health benefits of scaling up the best buy interventions. OHT is designed to strengthen health system analysis and costing and to develop financing scenarios at the country level. It is primarily designed to inform the development of national strategic health plans, by assessing parameters related to cost, impact and financing projections related to strengthening health systems and delivering costed and quantifiable strategic plans in low- and middle- income countries. Given its incorporation of epidemiological models that allow for prediction of health outcomes and costs in an integrated way across programmes and interventions, OHT has been used for previous global “investment cases”, including Cardiovascular disease 19, reproductive, maternal, newborn and child (RMNCH) health 20 and mental health21, as well as for the cost and impact analysis for the SDG Health Price Tag 15. The impact modules developed for cardiovascular disease, diabetes, asthma, COPD and cancers follow the structural format of the population models previously used in WHO’s cost-effectiveness programme, WHO- CHOICE 22,23.These are multi-state dynamic population life tables, taking account of competing risks amongst diseases, causes of death and interventions. The impact modules are populated at the regional level using the 21 Global Burden of Disease (GBD) regions, further analysed to develop complete epidemiological models. Data to populate the modules is derived from the 2010 GBD study along with the WHO Global Health Estimates database, and supplemented by literature reviews where data were missing. Risk factor modules impact the incidence of associated diseases using relative risks from the Global Burden of Disease Comparative Risk Assessment analysis24. Country-specific current risk factor prevalence was drawn from WHO Global Health Database, except for salt intake which was taken from the Global Burden of Disease estimates.25 A full list of epidemiological parameters used to populate the modules is available online at http://www.avenirhealth.org/software-onehealth.php , and is summarized in table 5. Table 5 Parameterization of the OneHealth Tool Data input Measured as Source Tobacco smoking Yes/No WHO GHDx (Country specific data for ages 20+ M/F)26 Alcohol intake Hazardous and WHO GHDx (Country specific data for ages 20+ M/F)26 harmful use Physical activity Insufficient vs WHO GHDx (Country specific data for ages 20+ M/F)26 sufficient Salt intake Grams per day Global Burden of Disease estimates 25 BMI BMI (kg/m2) WHO GHDx (Country specific data for ages 20+ M/F) 26 Systolic Blood Pressure mmHg (population WHO GHDx (Country specific data for ages 25+ M/F) 26 mean and SD) Cholesterol mmol/L (population WHO GHDx (Country specific data for ages 25+ M/F) 26 mean and SD) Stroke epidemiology Prevalence IHME GBD data27 Mortality RR mortality post stroke, adjusted by regional variations in stroke deaths28,29 Incidence DisMod 2 calculation using prevalence and mortality inputs30 IHD epidemiology Prevalence IHME GBD data27 10
Mortality RR mortality post IHD, adjusted by regional variations in IHD deaths 31,32 Incidence DisMod 2 calculation using prevalence and mortality inputs30 Diabetes epidemiology Prevalence IHME GBD data27 Mortality Incidence DisMod 2 calculation using prevalence and mortality inputs30 Asthma epidemiology Prevalence IHME GBD data27 Mortality Incidence DisMod 2 calculation using prevalence and mortality inputs30 COPD epidemiology Prevalence IHME GBD data27 Mortality Incidence DisMod 2 calculation using prevalence and mortality inputs30 Cervical Cancer Prevalence Globocan database33 epidemiology Mortality Incidence DisMod 2 calculation using prevalence and mortality inputs30 The impact modules are used to project two scenarios, one in which current intervention coverage is continued, and an alternative in which the intervention is scaled up at the rate as previously described, and with the health impact previously described. The difference between incidence and mortality rates, and healthy life years lived; between the two populations is the impact of the intervention. 11
ESTIMATING HEALTH CARE COSTS OF INTERVENTIONS The use of the OneHealth Tool to estimate the health costs for individual level interventions enables us to take an integrated approach to costs and benefits. Numbers of health services required are dynamically estimated over time, and are affected by population growth, mortality and disease incidence as interventions are scaled up. The costing analysis uses an ingredients approach which multiplies needs-based quantities by country-specific unit costs for intervention delivery. The prices used to develop unit costs come from the International Drug Price Indicator Guide, combined with the WHO-CHOICE price databases 34,35. The quantity assumptions used reflect current guidelines for intervention delivery and not necessarily delivery practices in countries. Estimates of the number of health services delivered by country and year enable calculation of the additional health systems capacity to deliver the interventions at scale. For the non-commodity intervention specific costs, the number of inpatient and outpatient visits is calculated in the OHT and then multiplied by the WHO-CHOICE unit prices for inpatient and outpatient visits36. Table 6 and 7 outline costing assumptions for policy/population wide interventions and individual pharmaceutical interventions, respectively. Table 6: Costing inputs for policy/population wide interventions Intervention Major costing assumptions Increase excise taxes and prices on tobacco products Taxation is considered a legislative intervention. Assumptions on human resource requirements are previously published37 Enact and enforce comprehensive bans on tobacco This is considered a legislative intervention. advertising, promotion and sponsorship Assumptions on human resource requirements are previously published37 Implement large graphic health warnings on all tobacco This is considered a legislative intervention. packages Assumptions on human resource requirements are previously published37 It is assumed that plain packaging would require additional legislation to large graphic health warnings. Additionally, implement plain/standardized packaging Eliminate exposure to second-hand tobacco smoke in all This is considered a legislative intervention. indoor workplaces, public places, public transport Assumptions on human resource requirements are previously published37 Implement effective mass media campaigns that educate Generic advocacy/awareness campaigns are included the public about the harms of smoking/tobacco use and as part of all health care interventions, with second hand smoke assumptions relying on information from the marketing literature. We estimated that 10 times the intensity would be required to enact behaviour change in line with previous costing estimates37 Reduce salt intake by engaging the industry in a Voluntary intervention assumptions are published in voluntary reformulation process previous work on the costs of scaling up NCD action37 Reduce salt intake through implementation of front-of- Legislative intervention assumptions are published in pack labelling previous work on the costs of scaling up NCD action37 Reduce salt intake through a behaviour change Behaviour Change communication is considered as communication mass media campaign an intensive mass media campaign. Costs have been developed based on literature reviews across public health and marketing to ensure adequate viewership 12
is reached Reduce salt intake through establishment of a Legislative intervention assumptions are published in supportive environment in public institutions such as previous work on the costs of scaling up NCD hospitals, schools and nursing homes to enable low action37 sodium meals to be provided Increase in excise taxes on alcoholic beverages Key categories of resource include human resources (e.g. administrators, lawyers), training (e.g. enforcement), meetings, mass media. Assumptions follow previous work on the costs of scaling up NCD action37 Enforcement of bans or comprehensive restrictions on Key categories of resource include human resources exposure to alcohol advertising, promotion and (e.g. administrators, lawyers), training, meetings, sponsorship (across multiple types of media) mass media. Assumptions follow previous work on the costs of scaling up NCD action37 Enforcement of restrictions on the physical availability Key categories of resource include human resources of retailed alcohol (via reduced density of retail outlets (e.g. administrators, lawyers), training, meetings, and reduced hours of sale mass media. Assumptions follow previous work on the costs of scaling up NCD action37 Behaviour Change communication is considered as Implement community wide public education and an intensive mass media campaign. Costs have been awareness campaign for physical activity which developed based on literature reviews across public includes a mass media campaign combined with other health and marketing to ensure adequate viewership community based education, motivational and is reached. The price was benchmarked against a environmental programs aimed at supporting successful BCC campaign for physical activity in behavioural change of physical activity levels. Australia 38 Table 7: Costing assumptions for pharmaceutical interventions Percent Number of Days Unit cost (USD) Cost per case receiving units/day per (2018) (USD) (2018) case Combination drug therapy for those post event or at 20% or greater risk. Drugs and supplies required per patient Hydrochlorothiazide, tablet, 25 95 1 365 0.0043 1.491025 mg Enalapril, tablet, 20 mg 40 1 365 0.01 1.46 Atenolol, tablets, 50 mg 25 1.5 365 0.01 1.36875 Amlodipine, tablet, 10 mg 40 0.5 365 0.08 5.84 Simvastatin, 15 mg 100 1 365 0.04 14.6 13
Laboratory tests per patient Blood glucose level test 30 1 1 2 0.6 Cholesterol test 30 1 1 2 0.6 Urine analysis 30 1 1 1.83 0.549 Urine sugar analysis 100 1 1 0.67 0.67 Facility visits per patient Outpatient 3 per year 97% of patients 7 per year 3% of patients The model calculates total costs for intervention implementation for each country. The sum of country level costs are then divided by the sum of the population from the countries in each income level to provide an average per capita cost for each income level. Costs are stratified by low income, lower-middle income and total costs. 14
ESTIMATING THE ECONOMIC AND SOCIAL RETURNS ON INVESTMENT TRANSLATING AVOIDED DEATHS INTO ECONOMIC RETURNS The economic modelling of mortality follows the cohort of avoided deaths for each of the years 2018 to 2030. Each cohort is classified by age and sex. The effect of avoided mortality on the labour force is calculated by taking the numbers of deaths avoided by age and sex and applying a corresponding labour force participation rate for this age, sex and year sourced from country specific International Labour Organization projections of labour force participation rates 39. The contribution that each of these labour force cohorts makes to economic output is calculated by multiplying the number in each age and sex category who participate in the workforce by a value of the GDP contribution per worker. To do this the average productivity is first calculated for 2018 by dividing the World Bank estimate of GDP in current US dollars by the labour force in that year. The total GDP generated is calculated by summing the GDP produced by each cohort for each year of the period in which they are in the labour force. TRANSLATING INCIDENT CASES AVOIDED INTO ECONOMIC RETURNS The contribution to GDP of each cohort of healthy persons is calculated in a similar way as for mortality using the same assumptions about participation rates and productivity in the cohort of avoided incident cases of CVD. Firstly, 11% of people who have a heart attack or stroke will leave the work force entirely. Those who avoid having an event will continue to participate in the workforce, this contribute the GDP per worker value each year. It is assumed that there is reduction in productivity of 0.5% due to absenteeism 40 (those who miss work days) and 3.7% due to presenteeism41 (those who attend work but are less productive than expected) for those with CVD who continue to work. These values are similar to those reported by Alsono et al 42and Bruffaerts et al 43 based on analysis of the World Mental Health Surveys. Again, the differential contribution to GDP from each cohort is summed to give a total contribution to GDP for each year. SOCIAL BENEFITS OF INCREASED YEARS OF HEALTHY LIFE It is common when estimating the benefits of improved health to put a value on being alive. The common term employed for this kind of statistic is the value of statistical life (VSL), but a more accurate term would be the value of (a small) risk reduction (VRR). Building on the results of Viscusi et al 44, Jamison et al 45 , estimated the value of a life year as between 1.4 and 4.2 times GDP per capita, averaging 1.6 globally. Stenberg et al 20 modified this approach by assuming that the value of a life year was 1.5 times GDP per capita and that the economic benefit was equal to GDP per capita, leaving a residual value of 0.5 times GDP per capita as the social benefit. Following this approach, we apply a value of 0.5 times GDP per capita to each healthy life year gained from the interventions to estimate the intrinsic value of longevity. An exception to this within our methodology is for cervical cancer where we were not able to capture incident cases avoided, and instead used a VSL of 1.5 times GDP per capita for each life year saved due to premature mortality avoided. CALCULATING THE RETURN ON INVESTMENT While the returns to an investment in health can be expressed using different but related metrics such as the internal rate of return, this paper uses benefit-cost ratios (BCR) to compare net present values of benefits and costs. These ratios were calculated by dividing the net present value (NPV) of the economic and social benefits from mortality and morbidity avoided by the net present value of the costs of the intervention. NPVs were calculated at a discount rate of 3% as is usual for analysis of health programs. 15
SUMMARY OF FINDINGS COSTS OF SCALING UP ACTION Lack of political prioritisation and action for non-communicable disease (NCD) prevention and control places the world at risk of failing to meet the SDG targets if action is not swift and ambitious. The NCD “best buy” interventions suggest high-priority, low cost interventions that can be implemented in all settings if political will can enable it. To rapidly implement all policy-level “best-buys”, and begin a concerted effort to scale up individual level “best-buys” to reach 50% coverage by 2030 will cost just $0.62 per capita in low income countries and $1.44 per capita in middle income countries – an average of just $1.27 per person, per country, per year (table 8). The majority of costs are for clinical services, with population level costs of policy interventions very low on a per capita basis (figure 2). Table 8: Investment needs for best-buys Investments needed Cost of implementation Total per Policy Total per Total per Total cumulative to Total cumulative to capita capita average capita 2030, $ millions 2023, $ millions average 2018-2023 max 2030 2024-2029 Total package of all 16 best-buys $45,296 $12,954 $0.55 $1.01 $1.27 Reduce Tobacco Use $2,485 $1,152 $0.05 $0.04 $0.04 Reduce the harmful use of Best-buys sub package alcohol $3,163 $1,495 $0.06 $0.05 $0.06 Reduce Unhealthy Diets $2,390 $1,256 $0.05 $0.04 $0.03 Reduce Physical Inactivity $159 $56 $0.00 $0.00 $0.00 Pharmaceutical management of CVD $32,109 $7,825 $0.33 $0.75 $1.00 Management of Cancer $5,424 $1,360 $0.06 $0.13 $0.14 16
Figure 2: Investment needs by intervention sub-package 17
HEALTH BENEFITS Current rates of non-communicable diseases are responsible for approximately 8.5 million premature deaths per year in low and lower middle income countries, with cardiovascular disease responsible for approximately 40% of these deaths, and cancer 27%. Implementing the best-buy interventions would prevent 8.1 million premature deaths by 2030 (table 7). This represents a reduction of almost 15% in total premature mortality due to NCDs, however when looking at CVD alone – which is the major health outcome of many of the best buys – the SDG target would be reached by 2028, and surpassed by 2030. Additional targeted interventions for cancer control and other NCD treatment are also needed to fully achieve the SDG target (Figure 3). Premature deaths avoided SDG target 2030 1,400,000 1,200,000 1,000,000 800,000 600,000 400,000 200,000 0 2018 2019 2020 2021 2022 2023 2024 2025 2026 2027 2028 2029 2030 Figure 3: Deaths avoided and SDG target for CVD mortality In addition to mortality prevention, these interventions will prevent many primary events from occurring, as many as 17 million stroke and ischemic heart disease events by 2030 (table 9). Consequently people will live longer, happier, healthier lives, with 72 million additional healthy life years lived. 18
Table 9: Health impact associated with scaling up best-buys Lives saved Healthy Life Years Gained IHD + Stroke Cases Avoided Total lives Total lives Total cases Total cases Policy Total HLY, Total HLY saved, saved, avoided, avoided, cumulative cumulative cumulative cumulative cumulative cumulative 2018 - 2023 2018 - 2030 2018 - 2023 2018 - 2030 2018 - 2023 2018 - 2030 Total package of all 16 best- 1,544,681 8,171,626 11,780,877 81,503,254 5,534,863 17,207,993 buys Reduce Tobacco Use 56,928 352,275 1,216,542 8,364,550 280,536 906,897 Reduce the harmful use Best-buys sub package 14,217 96,863 5,550,391 32,082,596 9,901 56,034 of alcohol Reduce Unhealthy 64,656 1,196,934 537,312 11,231,040 266,438 2,668,144 Diets Reduce Physical 1,218 10,173 11,584 110,831 8,043 29,628 Inactivity Pharmaceutic al management 1,147,735 5,560,231 2,607,893 20,593,805 839,478 3,938,440 of CVD Management of Cancer 140,047 670,346 - - - - 19
THE ECONOMIC BENEFITS OF INVESTMENT Investing in NCD prevention and control has not only health but economic benefits. People without NCDs work more days per year and work more productively. Preventing deaths from NCDs increases the volume of the workforce, also contributing to economic growth. Investing in the NCD best buys yields a return of at least $7 for every $1 invested by 2030. The economic benefits amount to $2.05 per person, per year on average by 2023, growing to an average of $9.03 per person per year between 2023 and 2029, and peaking $14.06 per person per year in 2030. RETURN ON INVESTMENT When considering that many of these interventions are preventive, and the full impact will be seen over a generation, this is a strong return over only 12 years (table 10). The highest return on investment is seen for investment in sodium reduction policies, which have a large health impact for a very low average cost. Table 10 : Economic benefits of best buys implementation Policy Low income countries Lower-middle income countries Overall Total package of all 16 best-buys $ 2.05 $ 8.01 $ 7.43 Reduce Tobacco Use2 $ 5.01 $ 7.98 $ 7.63 Reduce the harmful use of alcohol $ 3.45 $ 9.51 $ 9.13 Reduce Unhealthy Diets $ 5.61 $ 13.61 $ 12.82 Reduce Physical Inactivity $ 0.72 $ 3.28 $ 2.80 Best-buys sub package Pharmaceutical management of CVD $ 1.14 $ 3.54 $ 3.29 Management of Cancer $ 2.25 $ 2.76 $ 2.74 Due to the nature of these preventive interventions there is a differential timing between the costing and health and economic benefits. As demonstrated in figure 4 below, the economic benefits steadily rise over this 12 year period, not yet reaching the plateau. 2 Note that the impact of tobacco use on lung cancer is not included in the model, thus this should be considered as a minimum ROI 20
16.00 Costs 14.00 12.00 Economic benefits 10.00 8.00 6.00 4.00 2.00 - 2018 2020 2022 2024 2026 2028 2030 Figure 4: costs and economic benefits of the best-buy strategy 21
LIMITATIONS OF ANALYTICAL FRAMEWORK This modelling exercise should necessarily be considered normative and indicative at the global level. In order to fully grasp the costs and health benefits of scaling up action in countries, a country contextualisation process should be undertaken. The model used to underlie this analysis, the OneHealth Tool, is available free to download and use for countries to undertake this process. A contextualisation process involves reviewing the epidemiology taken from global databases (WHO, GBD) and comparing to local sources, identifying how interventions are delivered in countries and if these match the quantity assumptions used in this model, identifying how much is paid for different inputs into the interventions in the local setting – this may differ from global modelled databases – and using the countries’ realistic implementation plans. There are two main limitations which impact ROI values calculated within this analysis. Firstly, we are limited in the number of diseases for which we can prospectively model health impacts. This affects in particular tobacco, where cancers have not been modelled, however given the lag time between smoking cessation and reduction in cancer incidence this is unlikely to influence the ROI greatly over the 12 year period. This brings us to the second limitation which is the time frame through to 2030, the SDG target year. For many of the preventive interventions the full health benefit will not yet be realised. An excellent example of this is physical activity mass media campaigns. If extending the analysis through even 10 additional years, the health benefits increase by 2.5 times, which strongly influences the ROI. Further limitations related to the ROI calculation are associated with the use of the labor force participation rates from the ILO, which capture only the formal workforce. At the country level more may be known about the informal labour market in order to incorporate this into the analysis. Secondly, relying on GDP per capita and GDP per worker estimates necessarily produces lower economic benefits in countries of lower income. This leads to a correlation between income level and ROI in this type of analysis. This should not be interpreted as indicating that only higher income counties should invest in NCDs. As countries’ GDP increases over the coming years, higher productivity values will be seen in those countries currently classified as low income countries. Finally, the use of non-country-specific values for absenteeism and presenteeism has an unknown impact; however, the values used are quite low and we would thus anticipate represent a conservative approach. 22
REFERENCES 1. World Health Organization. ‘Best Buys’ and other recommended interventions for the prevention and control of noncommunicable diseases. Updated (2017) Appendix 3 of the Global Action Plan for the prevention and control of noncommunicable diseases 2013-2020; 2017. 2. International Agency for Research on Cancer. IARC Handbooks of Cancer Prevention. Tobacco Control. Effectiveness of tax and price policies for tobacco control; 2011. 3. Levy DT, Yuan Z, Luo Y, Mays D. Seven years of progress in tobacco control: an evaluation of the effect of nations meeting the highest level MPOWER measures between 2007 and 2014. 2016. 4. Allemandi L, Tiscornia MV, Ponce M, Castronuovo L, Dunford E, Schoj V. Sodium content in processed foods in Argentina: compliance with the national law. Cardiovasc Diagn Ther; 2015: 197-206. 5. Pietinen P, Valsta LM, Hirvonen T, Sinkko H. Labelling the salt content in foods: a useful tool in reducing sodium intake in Finland. Public Health Nutr 2008; 11(4): 335-40. 6. Law MR, Morris JK, Wald NJ. Use of blood pressure lowering drugs in the prevention of cardiovascular disease: meta-analysis of 147 randomised trials in the context of expectations from prospective epidemiological studies. 2009. 7. Do HT, Santos JA, Trieu K, et al. Effectiveness of a Communication for Behavioral Impact (COMBI) Intervention to Reduce Salt Intake in a Vietnamese Province Based on Estimations From Spot Urine Samples. J Clin Hypertens (Greenwich) 2016; 18(11): 1135-42. 8. Land M-A, Jeffery P, Webster J, et al. Protocol for the implementation and evaluation of a community- based intervention seeking to reduce dietary salt intake in Lithgow, Australia. BMC Public Health 2014; 14(1): 357. 9. The impact of school food standards on children's eating habits in England. 2017. http://www.cabi.org/cabebooks/ebook/20143140798. 10. Grimes CA, Campbell KJ, Riddell LJ, Nowson CA. Sources of sodium in Australian children's diets and the effect of the application of sodium targets to food products to reduce sodium intake. Br J Nutr 2011; 105(3): 468-77. 11. Taylor F, Huffman MD, Macedo AF, et al. Statins for the primary prevention of cardiovascular disease. Cochrane Database Syst Rev 2013; (1): Cd004816. 12. World Health Organization (WHO) and International Society of... : Journal of Hypertension. 2017. 13. Wald NJ, Law MR. A strategy to reduce cardiovascular disease by more than 80%. 2003. 14. Collaboration AT. Collaborative meta-analysis of randomised trials of antiplatelet therapy for prevention of death, myocardial infarction, and stroke in high risk patients. BMJ 2002; 324(7329): 71-86. 15. Stenberg K, Hanssen O, Edejer T, Bertram M, al e. Financing transformative health systems towards achievement of the health Sustainable Development Goals: a model for projected resource needs in 67 low- income and middle-income countries - The Lancet Global Health. Lancet Global Health 2017; 5(9): e875-87. 16. World Health Organization. WHO report on the global tobacco epidemic 2017. WHO 2017. 17. World Health Organization. Noncommunicable Diseases Progress Monitor 2017. WHO 2018. 18. Trieu K, Neal B, Hawkes C, et al. Salt Reduction Initiatives around the World – A Systematic Review of Progress towards the Global Target. PLoS One; 2015. 23
19. Bertram MY, Sweeny K, Lauer JA, et al. Investing in non-communicable diseases: an estimation of the return on investment for prevention andtreatment services. Lancet Forthcoming 2018. 20. Stenberg K, Axelson H, Sheehan P, et al. Advancing social and economic development by investing in women's and children's health: a new Global Investment Framework. Lancet 2014; 383(9925): 1333-54. 21. Chisholm D, Sweeny K, Sheehan P, et al. Scaling-up treatment of depression and anxiety: a global return on investment analysis. Lancet Psychiatry 2016; 3(5): 415-24. 22. Lauer JA, Röhrich K, Wirth H, Charette C, Gribble S, Murray CJ. PopMod: a longitudinal population model with two interacting disease states. Cost Eff Resour Alloc 2003; 1(1): 6. 23. Gopalappa C, Guo J, Meckoni P, et al. A two-step Markov processes approach for parameterization of cancer state-transition models for low- and middle- income countries. Medical Decision Making 2018; In Press. 24. GBD 2015 Risk Factors Collaborators. Global, regional, and national comparative risk assessment of 79 behavioural, environmental and occupational, and metabolic risks or clusters of risks, 1990–2015: a systematic analysis for the Global Burden of Disease Study 2015. The Lancet 2016; 388(10053): 1659-724. 25. Powles J, Fahimi S, Micha R, et al. Global, regional and national sodium intakes in 1990 and 2010: a systematic analysis of 24 h urinary sodium excretion and dietary surveys worldwide. BMJ Open 2013; 3(12). 26. World Health Organization. Global Health Observatory. http://gamapserver.who.int/gho/interactive_charts/mbd/life_expectancy/atlas.html. 27. Murray CJ, Vos T, Lozano R, et al. Disability-adjusted life years (DALYs) for 291 diseases and injuries in 21 regions, 1990-2010: a systematic analysis for the Global Burden of Disease Study 2010. Lancet 2012; 380(9859): 2197-223. 28. Katzenellenbogen JM, Vos T, Somerford P, Begg S, Semmens JB, Codde JP. Excess mortality rates for estimating the non-fatal burden of stroke in Western Australia: a data linkage study. Cerebrovasc Dis 2010; 30(1): 57-64. 29. Feigin VL, Forouzanfar MH, Krishnamurthi R, et al. Global and regional burden of stroke during 1990- 2010: findings from the Global Burden of Disease Study 2010. Lancet 2014; 383(9913): 245-54. 30. Barendregt JJ, Van Oortmarssen GJ, Vos T, Murray CJ. A generic model for the assessment of disease epidemiology: the computational basis of DisMod II. Popul Health Metr 2003; 1(1): 4. 31. Brønnum-Hansen H, Jørgensen T, Davidsen M, et al. Survival and cause of death after myocardial infarction: the Danish MONICA study. J Clin Epidemiol 2001; 54(12): 1244-50. 32. Moran AE, Forouzanfar MH, Roth GA, et al. The global burden of ischemic heart disease in 1990 and 2010: the Global Burden of Disease 2010 study. Circulation 2014; 129(14): 1493-501. 33. Globocan 2012 - Home. 2018. http://globocan.iarc.fr/Default.aspx (accessed Accessed September 2016. 34. Management Sciences for Health. International Drug Price Indicator Guidehttp://erc.msh.org/. 2010. 35. Bertram MY, Stenberg KE, Brindley C, et al. Disease control programme support costs: An update of WHO-CHOICE methodology, price databases and quantity assumptions. Cost-Effectiveness and Resource Allocation 2017; 15: 21. 36. Stenberg K, Fitzpatrick C, Lauer J. Facility visit unit costs. Cost effectiveness and Resource allocation Under review. 37. Chisholm D, Abagunde D, Mendis S. Scaling up action against noncommunicable diseases: how much will it cost? France: World Health Organization, 2011. 38. Leavy JE, Rosenberg M, Bauman AE, et al. Effects of Find Thirty every day®. http://dxdoiorg/101177/1090198112459515 2012. 39. International Labour Organization. ILO STAT. (accessed April 2015. 24
40. Mitchell RJ, Bates P. Measuring Health-Related Productivity Loss. Popul Health Manag; 2011: 93-8. 41. Wang PS, Beck A, Berglund P, et al. Chronic medical conditions and work performance in the health and work performance questionnaire calibration surveys. J Occup Environ Med 2003; 45(12): 1303-11. 42. Alonso J, Petukhova M, Vilagut G, et al. Days out of role due to common physical and mental conditions: results from the WHO World Mental Health surveys. Mol Psychiatry 2011; 16(12): 1234-46. 43. Bruffaerts R, Vilagut G, Demyttenaere K, et al. Role of common mental and physical disorders in partial disability around the world. Br J Psychiatry 2012; 200(6): 454-61. 44. Viscusi WK, Aldy J. The Value of a Statistical Life: A Critical Review of Market Estimates Throughout the World. Journal of Risk and Uncertainty 2003; 27(1): 5-76. 45. Jamison DT, Summers LH, Alleyne G, et al. Global health 2035: a world converging within a generation. Lancet 2013; 382(9908): 1898-955. 25
You can also read